摘要
由于脉冲神经元和突触复杂的时空动力学特性,训练脉冲神经网络比较困难,目前尚不存在公认的核心训练算法与技术。为此,提出一种基于直接反馈对齐(DFA)的精确脉冲时间(PREST-DFA)学习规则。受脉冲分层误差再分配(SLAYER)学习算法的启发,PREST-DFA使用基于脉冲卷积差的误差信号,输出层通过迭代方式计算出误差值,利用基于DFA的误差传输机制,将误差广播至隐藏层神经元,最后实现突触权值更新。仿真实验表明,实现了时间驱动的PREST-DFA学习算法具有精确脉冲时间学习能力。根据文献查询结果,这是首次验证基于DFA机制的学习算法可以在深层网络中控制脉冲的精确发放时间,说明DFA机制可以应用于基于脉冲时间的算法设计。另外还进行了学习性能和训练速度的比较,实验结果表明PREST-DFA能在较低的推理延迟下实现良好的学习性能,与采用相同学习规则使用反向传播训练的学习算法相比,能够加快训练速度。
Due to the complex spatiotemporal dynamics of spike neurons and synapses,training spike neural networks(SNNs)is relatively challenging,and there are currently no widely accepted core training algorithms and techniques.In this paper,we propose a learning rule with precise spike timing based on direct feedback alignment(PREST-DFA).Inspired by the learning algorithm called spike layer error reassignment(SLAYER),PREST-DFA uses error signals based on spike convolution differences.The output layer iteratively calculates the error values,and utilizes direct feedback alignment(DFA)to broadcast the error to hidden layer neurons,finally achieving synaptic weights update.We implement time-driven PREST-DFA,and simulation experiments demonstrate that PREST-DFA has precise spike timing learning capabilities and good biological plausibility.Based on literature search results,this is the first time to verify that learning algorithm based on DFA can control the precise fire time of spikes in deep networks,indicating that the DFA mechanism can be applied to algorithm design based on spike timing.We also compare learning performance and training speed.Experimental results show that PREST-DFA can achieve good learning performance with lower inference latency and can accelerate training speed compared to learning algorithms trained using backpropagation with the same learning rule.
作者
宁黎苗
王自铭
林志诚
彭舰
唐华锦
NING Limiao;WANG Ziming;LIN Zhicheng;PENG Jian;TANG Huajin(College of Computer Science,Sichuan University,Chengdu 610065,China;School of Computer Science and Technology,Zhejiang University,Hangzhou 310027,China)
出处
《计算机科学》
北大核心
2025年第3期260-267,共8页
Computer Science
关键词
脉冲神经网络
直接反馈对齐
学习规则
精确脉冲时间
在线学习
Spiking neural network
Direct feedback alignment
Learning rule
Precise spike timing
Online learning
